Last week, beginning ~10:00 25 October UTC and ending around 18:00 26 October UTC (a period of around 48 hours), we experienced several periods of significant API and email processing delays.
This instability was caused by an inefficient query pattern behind one of the conditions used in workflows. This inefficiency was revealed based on a usage pattern we had not seen before.
Once identified as the root cause, we took steps to mitigate this issue and introduce guardrails to prevent it arising in the future. We monitored for several days to ensure that all was fully resolved.
We have not seen the issue since and consider the issue resolved.
We are sorry for any inconvenience caused. If you have any follow up questions please contact us at firstname.lastname@example.org.